AI Agent for Chinese Chess (Xiangqi) - CS 221 Project
- logicblocker saint.seiya.0223@gmail.com | 2016 Autumn
- Stanford CS 221 Artificial Intelligence
# clone
git clone https://github.com/logicblocker/xiangqi-agent.git
cd xiangqi-agent
# Install dependencies
npm install
# start server
npm start
Open your browser, go to http://localhost:3000/ and you should be good to go
-
AI algofithms in the backend computation engine need to be enchanced (I was a bit hasty when then poster session was approaching >_<)
MCTSandTDLearningimplementations are not solid- reinforcement learning part seldom works
-
Effienciency need to be imporved
- Current computation engine is in javascript, which is not fast enough (I did not know how to call other executables from nodeJS then..., but now I do, please see below)
- You may implement a more powerful engine in C/C++ by calling a subprocess in server, NodeJS Child Process should help
Thanks for your interest :)



